Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/24374

Title: Multiple confidence intervals for selected parameters adjusted for the false coverage rate in monotone dose-response microarray experiments
Authors: Peng, Jianan
Liu, Wei
Bretz, Frank
Shkedy, Ziv
Issue Date: 2017
Publisher: WILEY
Citation: BIOMETRICAL JOURNAL, 59(4), p. 732-745
Abstract: Benjamini and Yekutieli () introduced the concept of the false coverage-statement rate (FCR) to account for selection when the confidence intervals (CIs) are constructed only for the selected parameters. Dose-response analysis in dose-response microarray experiments is conducted only for genes having monotone dose-response relationship, which is a selection problem. In this paper, we consider multiple CIs for the mean gene expression difference between the highest dose and control in monotone dose-response microarray experiments for selected parameters adjusted for the FCR. A simulation study is conducted to study the performance of the method proposed. The method is applied to a real dose-response microarray experiment with 16, 998 genes for illustration.
Notes: [Peng, Jianan] Acadia Univ, Dept Math & Stat, Wolfville, NS B4P 2R6, Canada. [Liu, Wei] Univ Southampton, S3RI, Southampton SO17 1BJ, Hants, England. [Liu, Wei] Univ Southampton, Sch Math, Southampton SO17 1BJ, Hants, England. [Bretz, Frank] Novartis Pharma AG, CH-4002 Basel, Switzerland. [Bretz, Frank] Shanghai Univ Finance & Econ, Sch Stat & Management, Shanghai, Peoples R China. [Shkedy, Ziv] Univ Hasselt, Ctr Stat CenStat, BioStat 1, Campus Diepenbeek,Agoralaan Gebouw D, B-3590 Diepenbeek, Belgium.
URI: http://hdl.handle.net/1942/24374
DOI: 10.1002/bimj.201500254
ISI #: 000404907800010
ISSN: 0323-3847
Category: A1
Type: Journal Contribution
Validation: ecoom, 2018
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
Published version485.69 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.